Project Summary/Abstract Accurate perception of others' emotion is critical for successful social interaction, yet it diminishes over the lifespan and is impaired across a wide range of disorders, including anxiety, depression, bipolar disorder, autism, and schizophrenia. The current proposal builds on a computational model developed by the PI, which posits that perceptions of social categories, including emotion categories, are driven not only by bottom-up cues inherent to targets (e.g., facial features) but also top-down factors harbored within perceivers, such as context and expectations. Although recent studies have documented the impact of such top-down factors in emotion perception, the mechanisms underlying these top-down impacts and how ?deeply? they manifest at lower levels of visual representation is far less clear. The overarching goal of this proposal is to use cutting- edge neuroimaging and behavioral methods to understand basic mechanisms of facial emotion perception, and particularly the impact by expectations and context. Functional magnetic resonance imaging (fMRI) will be used in conjunction with an innovative behavioral technique developed by the PI to accomplish 3 specific aims: (a) examine how context and expectations systematically bias the visual representation of a face's emotion, which is driven by OFC?FG interplay and depends on IFOF integrity; (b) test whether this top-down modulation is automatic and does not reflect some form of post-perceptual response bias; and (c) this mechanism this mechanism is shaped by inter-individual variability in emotion-conceptual knowledge, which provides a set of top-down expectations on the perception of facial emotion. The results will have numerous implications for the understanding of flexible emotion perception at the behavioral and neural levels, and will also help inform a variety of clinical processes involving the impairment of emotion perception and social dysfunction.